Visionds

[bwbelljr]

No description provided.

[eyalho]

Hey, have you trained enough to share results?
On MNIST, it scores badly. I changed epochs to 500, and retrain, hope to see and share results tomorrow..

[eyalho]

Why epoch=5?
All other yml used 500.
It give bad results training "as is":
{'ACC': 0.1135, 'ARI': 0.0, 'NMI': 6.755387526067596e-16, 'ACC Top-5': 0.5134, 'hungarian_match': [(0, 0), (1, 9), (2, 2), (3, 3), (4, 4), (5, 5), (6, 6), (7, 7), (8, 8), (9, 1)]}
Did you achieve good result in one of experiments?

[eyalho]

better live unchanged

[wvangansbeke]

Hi @bwbelljr @eyalho,

Your contribution is very much appreciated, @bwbelljr!

I haven't found the time to check this pull request yet, but it looks useful. I want to make sure that it works before merging it.
Augmentations and other hyperparameters might be different for MNIST @eyalho.